INTRODUCTION

Suppose an organism sees that a tiger is at hand. The organism must decide whether this tiger is dangerous. Two types of strategy are available for making this decision.

An individual who learns will believe the proposition, or believe its negation, depending on the character of its experiences. For example, the organism might attend to whether the tiger is wagging its tail, and decide what to believe on that basis. The alternative strategy is for the organism to decide on the basis of a priori prejudice. It will believe that the tiger is dangerous (or that it is not) irrespective of the character of its experience.

Learning is a conditional strategy, whereas adhering to an a priori prejudice is unconditional. The prejudiced individual conforms to the rule always believe that tigers are dangerous. The learner conforms to the rule believe that a tiger is dangerous if your experience has characteristic C, but believe that the tiger is not dangerous if your experience has characteristic D.

The problem to be investigated here concerns the adaptive advantage of learning and a priori prejudice. Under what circumstances is learning advantageous? This problem is subsumable under the more general heading of determining when an obligate response is fitter than a facultative response.